Abstract:We propose a strategy to compress and store large volumes of scientific data represented on unstructured grids. Approaches utilizing tensor decompositions for data compression have already been proposed. Here, data on a structured grid is stored as a tensor which is then subjected to appropriate decomposition in suitable tensor formats. Such decompositions are based on generalization of singular value decomposition to tensors and capture essential features in the data with storage cost lower by orders of magnitude. However, tensor based data compression is limited by the fact that one can only consider scientific data represented on structured grids. In case of data on unstructured meshes, we propose to consider data as realizations of a function that is based on functional view of the tensor thus avoiding such limitations. The key is to efficiently estimate the parameters of the function whose complexity is small compared to the cardinality of the dataset (otherwise there is no compression). Here, we introduce the set of functional sparse Tucker tensors and propose a method to construct approximation in this set such that the resulting compact functional tensor can be rapidly evaluated to recover the original data. The compression procedure consists of three steps. In the first step, we consider a fraction of the original dataset for interpolation on a structured grid followed by sequentially truncated higher order singular value decomposition to get a compressed version of the interpolated <a class="link-external link-http" href="http://data.We" rel="external noopener nofollow">this http URL</a> then fit singular vectors on a set of functional basis using sparse approximation to obtain corresponding functional sparse Tucker tensor representation. Finally, we re-evaluate the coefficients of this functional tensor using randomized least squares at a reduced computational complexity. This strategy leads to compression ratio of orders of magnitude on combustion simulation datasets.

Tensor-Train Compression of Discrete Element Method Simulation Data

Hardware-Enabled Efficient Data Processing with Tensor-Train Decomposition

Distributed Non-Negative Tensor Train Decomposition

TetraFEM: Numerical Solution of Partial Differential Equations Using Tensor Train Finite Element Method

Randomized Functional Sparse Tucker Tensor for Compression and Fast Visualization of Scientific Data

An Incremental Tensor Train Decomposition Algorithm

Use of Tensor-Train Decompositions with a Discrete Velocity Boltzmann Solver

Hybrid Tensor Decomposition in Neural Network Compression

Nonlinear Tensor Train Format for Deep Neural Network Compression

Compressing 3DCNNs based on tensor train decomposition

TensorCodec: Compact Lossy Compression of Tensors without Strong Data Assumptions

High-Performance Tensor-Train Primitives Using GPU Tensor Cores

Efficient Computation of Tucker Decomposition for Streaming Scientific Data Compression

Tensor-Train Point Cloud Compression and Efficient Approximate Nearest-Neighbor Search

Tensor Networks for Big Data Analytics and Large-Scale Optimization Problems

Parallel algorithms for computing the tensor-train decomposition

A Distributed Tensor-Train Decomposition Method for Cyber-Physical-Social Services.

Scalable Tensor-Train-Based Tensor Computations for Cyber-Physical-Social Big Data

Scalable Hybrid Learning Techniques for Scientific Data Compression

Efficient Leverage Score Sampling for Tensor Train Decomposition

A Robust Low-Rank Tensor Decomposition and Quantization Based Compression Method